| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
TRANSLATIONAL PHYSIOLOGY
Department of Biomedical Engineering, Marquette University, Milwaukee, Wisconsin 53201-1881
Submitted 14 July 2003; accepted in final form 22 March 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). Locomotor pathways receive strong input from hip proprioceptive cues (Grillner and Rossignol 1978; Hiebert et al. 1996; Kriellaars et al. 1994; Lam and Pearson 2001; Pang and Yang 2000, 2001; Van de Crommert et al. 1998), which is consistent with triggering of extensor spasms by imposed extension movements of the hip (Schmit and Benz 2002). This study will show that imposed sinusoidal hip movements entrain reflex activity in leg muscle groups in a manner that is consistent with locomotion.
Individuals with chronic SCI often have a secondary condition referred to as spasticity. Spasticity is classically defined as a motor disorder characterized by a velocity-dependent increase in tonic stretch reflexes with exaggerated tendon jerks, resulting from hyperexcitability of the stretch reflex (Lance 1980). While this definition may be appropriate for many populations with neural injuries, it does not encompass many of the clinically observed components of spasticity in SCI. One specific component of SCI spasticity that is poorly described by this definition is the motor behavior often named "extensor spasms."
Extensor spasms, which result in substantial reflexive muscle activity, are commonly observed in the clinic during movement of SCI patients from a sit to a supine position (Kuhn 1950; Little et al. 1989). The sit to supine movement consists largely of a bilateral hip extension, which has lead to experimental reproduction of spasms using controlled unilateral extension of the hip (Schmit and Benz 2002). The response to this trigger consists of a multijoint reflex and is characterized by hip flexion, knee extension, and ankle extension torques. Since this is a multijoint response, it cannot be attributed to stretch reflexes alone; therefore extensor spasms must be mediated through polysynaptic pathways involving activation of organized interneuronal circuits located within the isolated spinal cord that are modulated by hip proprioceptive input.
Because of the dependence on hip afferents, it has been postulated that the multijoint extensor reflex response is associated with interneuronal pathways that make up spinal locomotor circuits. Adults with chronic SCI have shown evidence for hip afferent modulation of locomotion during partial weight supported treadmill training, specifically associated with enhanced swing following pronounced imposed hip extension at the end of stance (Dietz et al. 1998, 2002; Dobkin et al. 1995; Harkema 2001). Locomotion is also modulated by hip proprioceptors in human infants walking on a treadmill, again with swing enhanced by exaggerated hip extension during terminal stance (Pang and Yang 2000). These observations are consistent with animal experiments in which the swing phase of the locomotor cycle is initiated and modulated by the hip afferent input from hip flexor muscles in the spinalized cat (Grillner and Rossignol 1978).
Entrainment of alternating flexor and extensor activity also occurs with the mechanical application of a sinusoidal movement pattern at the hip of the spinalized cat. The entrainment pattern is consistent with the frequency of the sinusoidal oscillation (Andersson and Grillner 1983; Conway et al. 1987; Kriellaars et al. 1994). Previous tests of hip-triggered reflexes in human SCI show a strong prolonged response when the limb is moved into extension and held; however, significant cocontraction is observed, raising questions about whether the characteristic torque response of hip flexion, knee extension, and ankle extension is the result of global muscle activation, with the strongest muscle groups having the dominant torque response (Schmit and Benz 2002).
In this study, we tested whether sinusoidal oscillations of the hip in chronic human SCI would produce entrainment of muscle activity of the leg, similar to studies in spinalized cats. Power spectral analyses of both torque and smoothed EMG responses were conducted to examine entrainment. Resultant vector analysis in a polar coordinate system was used to determine the relative phasing of leg muscle activity during the imposed movements. The sinusoidal responses were contrasted to the response to ramp and hold perturbations and compared with typical locomotor patterns to identify the similarities and differences with walking. We postulated that similarities to locomotion would exist due to the excitation of common neural pathways in normal walking and during extensor spasms in chronic SCI.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Ten subjects with chronic SCI were recruited into this study. Inclusion criteria included a history of SCI (>4 mo) with associated spasticity. Participants (mean age, 30.7 ± 6.94 yr; range, 19–42 yr) included four clinically complete [American Spinal Injury Association (ASIA) classification A] and six clinically incomplete (ASIA B or C) individuals with cervical (6 or 7 subjects) or thoracic (3 or 4 subjects) SCI (Table 1). At the time of the study, 5 of the 10 subjects were prescribed antispastic medications to reduce the intensity and frequency of spasms. Exclusionary criteria included multiple CNS lesion sites or secondary lesions of the cord, the presence of significant complications such as skin breakdown, urinary tract infection, other secondary infections, heterotopic calcification, respiratory failure or other concurrent illness limiting the capacity to conform with study requirements, the inability to give informed consent, and significant osteoporosis. Informed consent was obtained, and all procedures were conducted in accord with the Helsinki Declaration of 1975 and approved by the Institutional Review Board of Marquette University.
|
A novel apparatus, shown in Fig. 1, was constructed for measuring the multijoint torque response to imposed movements at the right hip. This apparatus included a knee-ankle brace with built-in torque transducers aligned with the axes of rotation of the knee and ankle. A footplate included a clamp to be placed on the dorsum of the foot and a strap to secure the heel. The hip-knee and knee-ankle links were adjustable to fit a wide range of leg sizes. The entire leg brace was affixed to a newly constructed velocity controlled servomotor drive system (MT704A1-R1C1, Kollmorgen, Northampton, MA). Hip torque, knee torque, and ankle torque were measured using hollow-flanged transducers (Himmelstein, Hoffman Estates, IL). Position of the hip joint was measured using a potentiometer coupled to the servomotor drive shaft.
|
All signals were low-pass filtered (450 Hz), and sampled at 1,000 Hz using a data acquisition card (National Instruments, Austin TX) on a personal computer. Custom LabVIEW software (National Instruments) was used for acquiring the data as well as outputting the velocity command signal for the servomotor system.
Imposed hip movements
Movements were imposed to the right hip of all 10 subjects. Each subject was transferred to a tri-section therapy table and placed in a supine position. The center of rotation of the right hip joint was aligned with the axis of rotation of the servomotor system, and the brace was adjusted to align the ankle and knee with the appropriate torque transducer. Alignment of the hip was confirmed by a lack of leg translation during imposed flexion and extension of the hip. The pelvis was secured to the table with a strap across the iliac crest to inhibit pelvic rotation. The foot was placed in a footplate and secured using a clamp placed on the dorsum of the foot and the heel. The leg was placed in the brace with the knee at 10–30° flexion and the ankle at 20–25° plantarflexion, as summarized by Table 2. The initial knee angle was set to 30° flexion. Manual hip extension perturbations were made to elicit an extensor spasm response. If no response occurred, the knee was extended further, which usually helped in eliciting a larger response. After the ankle and knee angles were set, these joints were held isometric for the duration of the test. The contralateral limb was supported in a slightly flexed position at the hip.
|
0.2, 0.3, and 0.35 Hz) with no pause in extension or flexion. These frequencies roughly correspond to stepping speeds of 0.21, 0.31, and 0.36 m/s, respectively. Table 2 displays the peak velocities as they corresponded to the different subject range of motions. Note that the peak velocity varied between subjects since the frequency was controlled and the hip range of motion was subject-dependent. A timed delay of 3 min was allowed between trials. The protocol began with a ramp stretch movement followed by three sinusoidal movements with the three frequencies applied in random order. This format was repeated for a total of three ramp stretch movements and nine sinusoidal movements. At the end of the 12 trials, the subject's leg was moved slowly in extension at 2° increments to measure the passive torque of the hip. Also, the leg was oscillated from 40° flexion to 20° flexion at frequencies of 3 and 5 rad/s to identify the inertial properties of the leg.
|
Ramp stretch movements. Joint torque data were obtained for the hip, knee, and ankle during the 10-s hold periods, with the hip in the maximum extended or flexed position. The gravitational/passive torque offset was removed by subtracting a torque measurement in the end position, when there was no muscle activity (e.g., at the end of the hold period). The signals were low-pass filtered (LPF, 5 Hz) using a fourth order Butterworth filter (butter/filtfilt; Matlab command, The Math Works, Natick, MA) and plotted against time. The pattern of muscle activity was identified for each subject and related to the measured reflex torque. The duration of the torque response was calculated using a threshold of ±3 Nm for the hip, ±1 Nm for the knee, and ±0.5 Nm for the ankle. These values were used because they eliminated inclusion of signal noise related to resonant vibration during the isometric reading. The resulting values represented the reflex torque at each joint and were used as a measure of the net reflex response at each joint. Rectified EMGs were evaluated to detect the timing of muscle activity during and following imposed hip movements.
Sinusoidal movements. Torque data for the hip, knee, and ankle were acquired during the entire movement cycle. To calculate the reflexive hip torque, the gravitational torque, passive resistance, and inertial torque were removed. The effects of gravity, passive joint resistance, and inertia were each calculated using a separate set of hip perturbations. Once determined, the net reflex response was calculated by subtracting the inertia, gravity, and passive torques from the torque measured during the oscillations.
Passive resistance and gravitational torque of the leg were determined by moving the leg throughout the entire range of motion at 2° increments, pausing for 2–10 s for a total of 20 increments throughout the range of motion. The mean torque signal was calculated during each pause and the resulting signal contained only the passive and gravitational torque. Gravitational torque alone was calculated from the mean torque measurements collected in the middle of the range of motion, during which time the passive resistance was negligible. A cosine function, multiplied by a scalar (Kleg) was fit to the torque data using a least squares regression (backslash operator; Matlab command), resulting in an estimate of the gravitational torque (Eq. 1)
 | (1) |
The passive resistance of the hip joint was then calculated by fitting a third order polynomial (polyfit/polyval; Matlab command) to the mean torque data, with the gravitational torque subtracted (Eq. 2)
 | (2) |
The inertial properties of the leg were estimated from hip torque data obtained during separate oscillations of the leg from 40° flexion to 20° flexion at 3 and 5 rad/s. By keeping the hip in a flexed position, no muscle activity was elicited, and no passive resistance was encountered; therefore the inertial constant, Ileg, could be calculated by correcting for gravity only (Eq. 3, a, b, and c). Ileg was determined through least squares regression (backslash operator; Matlab command)
 | (3a) |
 | (3b) |
 | (3c) |
From the gravitational constant Kleg, the inertial constant Ileg, and the third order polynomial coefficients associated with the passive resistance, the active torque was calculated for each trial using Eq. 4
 |
 | (4) |
The active torques at the knee and the ankle were calculated by correcting only for gravity and inertial artifacts of the foot-shank and foot, respectively. The passive resistance of the knee due to biarticular muscles of the hip and knee were assumed to be negligible (<2 Nm), a result supported by subjective observations of knee torque during slow trials into hip extension. The resulting torque measurements represented the reflex response at each joint. These biomechanical measurements of the responses were absolute measures of the leg output, which could be compared across subjects to determine the relative size of the response. The disadvantage of the torque measurements is that they could not account for muscle coactivation at the joints, which required EMG measurements for interpretation.
For analysis, smoothed rectified EMG signals were calculated. Surface EMGs were rectified and enveloped using a 4-Hz, low-pass, eighth order Butterworth digital filter (butter/filtfilt; Matlab command). EMG signals were then summed over the last five cycles, normalized to the maximum, and plotted in a polar coordinate system, where 180° represents full extension of the leg. The last five cycles were used in this analysis to eliminate the effects of transient modulations in the total responses, which were occasionally observed in the first five cycles.
Frequency analysis. To determine whether oscillation of the hip produced reflex responses that were entrained to the movement, the frequency content of the reflex response was identified. Both smoothed rectified EMG and torque signals were examined for spectral content using a Discrete Fourier Transform (DFT) method that squares the magnitude of the DFT within Matlab (periodogram; Matlab command). All data were zero-padded to length of 217 (131,072). A rectangular window was used to obtain the best resolution. The maximum spectral peak was obtained from the power spectral density estimate (PSD). The maximum spectral peaks were examined and compared with their corresponding time series signal to determine whether the peak occurred at the frequency of the movement. Similar peak frequency content was used to indicate entrainment of the response to the movement.
Phase analysis.
Muscle timing patterns during each sinusoidal cycle were examined using circular statistics according to methods outlined by Batschelet (1981). Phase analysis of EMG signals was only pursued if the EMG signals showed sufficient muscle activity. Amplified EMG noise levels typically did not exceed 5 mV; therefore a threshold of 10 mV for enveloped and rectified EMG was utilized as the criterion for including individual trials in the analysis. Similarly, a threshold of ±5 Nm for hip extension/flexion, ±4 Nm for knee extension/flexion, and ±1 Nm for ankle extension/flexion was implemented for inclusion of reflex torque signals. This threshold was set so as not to include noise associated with the sinusoidal torque calculation of active muscle torque.
Phase analysis consisted of using rectified and smoothed EMG signals that were normalized from 0 to 360, where 180 represents full extension of the hip. The signals were plotted in a polar coordinate system. For each movement cycle, the Cartesion coordinates of the resultant vector for each muscle were calculated using Eq. 5, a and b. From these x,y coordinates, the polar angle was found using Eq. 5c. The vector length, r, was normalized to a unit vector size in a conversion back to x,y coordinates. After normalization, the mean polar angles and vector length, r, were calculated (Eq. 5, c and d) across trials of similar frequency if the criteria for a minimum response were met. The mean polar angles and vector lengths were then used for a phase analysis. Similar analysis was performed on joint torque signals from sinusoidal trials. The torque signals were first half-wave rectified in both flexion and extension, and then analyzed to find the mean polar angle of the resultant vector using the aforementioned procedure.
To determine whether there was a significant phasing in EMG and torque signals, Raleigh's test for one-sidedness was performed (
= 0.05). All data sets that showed a significant trend were plotted in polar coordinates, where the vector length, r, was used to determine significance (Batschelet 1981)
 | (5a) |
 | (5b) |
 | (5c) |
 | (5d) |
Amplitude analysis. The peak hip flexion, knee flexion, ankle flexion, hip extension, knee extension, and ankle extension torques were calculated for each movement trial in each subject to identify the effect of movement frequency on the amplitude of the response. The mean of the three 1.2-, 1.88-, and 2.2-rad/s trials were calculated for each subject. A two-factor ANOVA was used compare the effect of movement frequency and subject on the mean torque for each joint to determine whether the magnitude of the reflex response was velocity dependent.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
In this study, EMG and reflex torque responses were measured during both ramp-hold and sinusoidal hip movements to show that sensory input from the hip joint is important in modulating the timing and duration of leg muscle activity in chronic human SCI (n = 10). Typical EMG and torque responses for both movement types are shown in Fig. 3. Note that EMG responses occasionally exhibited bursting activity, consistent with clonus (Beres-Jones et al. 2003). In the EMG analysis, the signals were rectified and low-pass filtered at 4 Hz, effectively filtering out this clonic response.
|
Typically, the duration of the reflex response to hip extension was significantly reduced when a flexion was immediately imposed to the hip. In Fig. 3, it can be observed that both the EMG and torque responses were prolonged when holding the leg in extension. In fact, 7 of the 10 subjects with hip flexion responses produced a hip flexion torque of >3 Nm for longer than 5 s. In addition, four of five subjects with knee extension responses held knee extension torques above 3 Nm for over 4 s. Ankle extension responses were held above 1 Nm for more than 3 s in three of six subjects that exhibited ankle extensor torques. In four subjects, hip flexion moments were held above 3 Nm for the entire 10 s that the limb was held in extension. These prolonged responses ceased when the limb was moved into flexion. This cessation of hip flexion and knee extension torques on flexion of the limb were consistent with the results seen with the sinusoidal perturbations, in which the duration of both the EMG and torque response were equal to the period of the imposed sinusoidal oscillation (see Torque and EMG during sinusoidal oscillations).
Torque and EMG during sinusoidal oscillations
The duration of the torque and EMG responses during sinusoidal oscillations was dependent on the frequency of the movement. Typically, the EMG and torque responses were entrained to the frequency of the oscillation. These observations were quantified by identifying the frequencies of the peaks of the power spectrum of each signal.
In general, the maximum spectral peak for the torque response occurred at the frequency of the movement in all 10 subjects; however, there were exceptions for individual trials in subjects 1, 4, 5, 6, 7, 9, and 10. Figure 4 shows the power spectrum for hip, knee, and ankle torque responses during a single trial in subject 7. Peak A corresponds to the frequency of the movement. In Fig. 4, this peak was maximal in hip and knee torque responses. Peak A was the maximum spectral peak in 88 of 90 hip, 90 of 90 knee, and 59 of 90 ankle movement trials exhibiting entrainment. The number 90 represents the total number of trials across all subjects and frequencies (3 frequencies x 3 trials x 10 subjects). Peak B (Fig. 4) corresponds to twice the frequency of the movement. This peak was occasionally the largest (ankle responses, 4 of 90; hip, 2 of 90; knee, 0 of 90). Finally, peak C corresponds to a lower frequency component. This phenomena was observed at the ankle joint in 27 of 90 trials. Figure 4 shows that this low-frequency response corresponded to a large-amplitude, long-duration ankle plantar flexion (extension) that increased during the first three cycles. It was also observed that the high-frequency peak (peak B) was larger than the power at the oscillation frequency (peak A) in a majority of the cases where peak C was the maximum. The torque versus hip angle plots in Fig. 4 show that, in cases where a large high-frequency peak (peak B) was observed, the ankle created an extension moment on both flexion and extension of the hip, accounting for the location of the spectral peak at double the oscillation frequency.
|
EMG and torque phase analysis in sinusoidal oscillations
Muscle activity was synchronized with movement in cases where significant EMG signal was recorded. Surface EMG recordings demonstrated unambiguous muscle activity in 8 of 10 subjects (smooth rectified activity >10 mV). For example, EMG from the Sol, VM, RF, and MH for five hip movement cycles is shown in Fig. 5. The last five cycles were used for additional analysis because there were some cycle-to-cycle changes in the reflex torque response over the first five cycles. These changes are shown in hip, knee, and ankle torque responses in Fig. 4, A, D, and G. The EMG responses were rectified, smoothed, and plotted on a polar plot, as shown in Fig. 6A.
|
|
There were two exceptions to the muscle activity pattern shown in Fig. 6A. Figure 6B is an example of one exception, from subject 10. In this case, the VM and MH were in phase with each other during the flexion phase of the cycle. In addition, VM and RF activity had peaks just before peak extension occurred. The RF also had a peak just after full extension. The RF activity began to subside after the MH began to fire. Both subjects 6 and 10 showed this response; however, the response was only observed occasionally (subject 6: 7 of 9 trials; subject 10: 5 of 9 trials). The VM was out of phase with the MH in the remainder of the trials.
The polar angles of the mean resultant vector for the EMG signals, which were used determine the overall phasing each muscle, also showed a phasing of the RF, VM, and Sol activity with hip extension and MH phasing with hip flexion. Figure 7 shows the angular locations of the resultant vectors, for the 2.2-, 1.88-, and 1.2-rad/s movement frequencies, for trials that met Raleigh's test criteria (P < 0.05; see METHODS for other exclusion criteria). A mean resultant vector phase angle of 180° corresponds to muscle timing in phase with hip extension movements. The phase plots show that, for the majority of trials, the RF, VM, and Sol are in phase with the hip extension movement. The majority of MH activity is in phase with the hip flexion movements with only a slight phase lead. As Fig. 7 shows, TA, MG, and Add muscle activity did show patterns that were consistent across subjects. The number of subjects that yielded significant patterns for one-sidedness was especially small in the MG and TA.
|
|
There was a large disparity in the magnitude of the joint torque responses across subjects. Figure 9 shows the mean peak torque for hip, knee, and ankle torques for each frequency of the oscillation for all 10 subjects. Positive torque represents extension and negative values correspond to flexion responses. From Fig. 9, subjects 4, 5, 6, 7, 9, and 10 yielded the largest amplitude responses regardless of joint or direction.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The results from imposed hip oscillations contrasted with the responses to ramp and hold perturbations of the hip. Ramp and hold movements produced reflex activity similar to previously reported results (Schmit and Benz 2002), consisting largely of hip flexion, knee extension, and ankle extension torques in response to hip extension and predominant co-activation of the ankle and knee musculature, as reflected in the EMG data. These results were similar to patient descriptions of extensor spasms. Conversely, the organization of the muscle activity patterns during hip oscillation appeared to depend on continued movement of the hip.
Role of interneurons in the extensor spasm response
Spastic reflex behaviors, which have traditionally been attributed to velocity-dependent homonymous stretch reflexes (Ashworth 1964; Lance 1980), include hyperexcitable multijoint reflexes in chronic human SCI. The traditional definition of spasticity as a hyperexcitable stretch reflex has proven to be adequate when describing spasticity in many types of CNS disorders such as stroke (Schmit 2001), cerebral palsy (Engsberg et al. 2000), and multiple sclerosis (Sinkjaer et al. 1993). In contrast, SCI individuals also show other spastic reflexes, such as flexor spasms, which have been associated with an increased flexor reflex response to skin stimuli (Dimitrejevic and Nathan 1968; Shahani and Young 1971) or movement of the ankle (Schmit et al. 2000). Similarly, coactivation of muscles across several joints has been observed in response to a number of stimuli (Beres-Jones et al. 2003; Dimitrejevic and Nathan 1967a,b), including imposed hip extension movements (Schmit and Benz 2002). The results from this study are consistent with multijoint reflex activation playing a role in spasticity. Clearly movement at the hip produces reflex activity of the ankle and knee musculature that is not stretched by the movement (e.g., Fig. 3). These observations suggest that spastic reflexes in human SCI involve more than homonymous stretch reflex pathways, implicating interneuronal pathways that have the potential to coordinate the reflex response.
The increased excitability of reflex pathways involving spinal interneurons raises the question of whether single joint movements can trigger organized multijoint responses that have a functional correlate. Similar to previous studies, the ramp and hold stretch responses in this study show that single joint movements can trigger multijoint responses, including coactivation at the ankle and knee. In addition, peak joint torque was found to increase with increases in speed during hip oscillations (Fig. 10), suggesting that the response may be mediated by spindle afferents. In particular, this mechanism may account for activity of the hamstrings, which are stretched (and activated) during hip flexion. In addition to these results, however, we observed that the EMGs and joint torques became entrained to oscillatory movements (Figs. 4–8), with different muscles showing activity in separate phases of the oscillation. In addition, the entrainment occurred in muscle groups that do not cross the hip joint. The entrainment of the activity alone suggests that the reflex produces more than coactivation at the joints, raising the question of whether the interneuronal pathways that are activated in this reflex correspond to some motor function. Locomotor pathways are a potential neural substrate of these reflex responses to hip oscillation.
Are these interneuronal pathways associated with the spinal centers for locomotion?
The initiation of the organized responses observed in this current study is likely to be due to hip afferent feedback to the spinal cord as evidenced by the entrainment of the muscle activity (Figs. 4, 5, and 7). Hip afferents have also been shown to have important roles in the modulation of locomotion. The spinalized cat model has revealed that the swing phase of locomotion is initiated by extension of the hip (Grillner and Rossignol 1978). In addition, transitions from stance to swing in the human infant are modulated by hip joint afferents (Pang and Yang 2000, 2001). Our results show that during an imposed extension of the hip, the hip produces a flexion moment (Fig. 8), which is a muscle pattern consistent with the initiation of swing.
Fictive locomotion is influenced by hip afferents in a similar manner. Imposed sinusoidal oscillation to the hip joint during fictive locomotion of the immobilized spinal (Andersson and Grillner 1981, 1983; Conway et al. 1987) and decerebrate cat (Kriellaars et al. 1994) entrain alternating flexor and extensor muscle activity to the frequency of the movement. The torque and EMG responses to imposed hip oscillations in this study were also entrained to the frequency of the movement, as shown by dominant power in the EMG and torque signals at the frequency of the movement (Fig. 4), suggesting a neural mechanism similar to the one implicated during fictive locomotion in the cat. In contrast, the duration of the ramp and hold responses lasted several seconds, and cocontraction of the hip, knee, and ankle musculature was more prominent than cocontraction during sinusoidal responses.
Hip afferent input has also been shown to be one factor that is important in the modulation of alternating flexor and extensor rhythms in adult humans with chronic SCI. Involuntary locomotor-like activity was found in a single incomplete SCI subject (Calancie et al. 1994). This activity was induced when moving from a sit to supine position, with the knees fully extended. This activity suggests that the spinal center for locomotion is modulated by hip afferent input. Other studies have provided anecdotal evidence of alternating flexor and extensor activity after sit to supine movements (Kuhn 1950). These observations are consistent with modulation of locomotor rhythms through hip afferent input during body weight supported treadmill training of patients with chronic SCI (Dobkin et al. 1995) as well as with the observed modulation of muscle activity with hip oscillation in this study.
Many studies have been conducted that show that reflex pathways associated with locomotion are recruited using a treadmill, partial body weight support, and the assistance of physical therapists moving the legs in a gait pattern (Dietz 2002; Dietz et al. 1995, 1998; Dobkin et al. 1995; Harkema et al. 1997). During these activities, it has been noted that extension of the limb produces an involuntary hip flexion that initiates swing (Dobkin et al. 1995). These results are consistent with the hip torque produced during this experiment; however, the knee extension torque observed with imposed hip extension was inconsistent with swing, rather, it is typically more closely associated late stance phase. Figure 11 outlines the EMG activity differences between non-SCI gait (Perry 1992), SCI gait (Dobkin et al. 1995; Pepin et al. 2003), and the unilateral sinusoidal oscillations in this study. There are clear similarities in the EMG patterns for the MH, Sol, and RF. The quadriceps (VM/ VL) activity in all cases was, however, much more prolonged and in most cases completely out of phase with what is seen in non-SCI gait. A possible reason for this difference in activity pattern is the absence of other afferent input to the spinal locomotor centers.
|
The spontaneous alternating flexion and extension movements triggered by sit to supine movements discussed earlier (Calancie et al. 1994) were manifested only if the hip extension occurred while the knees were extended. In this study, reflex responses would not occur if the knees were flexed beyond a critical angle. For example, responses typically occurred if the knee was flexed to about 30°; however, in some cases the knee had to be extended further to elicit a response. In one case the knee was extended to 10° flexion to get a response. This is evidence that knee proprioceptors may play a role in the modulation of this activity. The knee proprioceptors have not been shown to play a large role in modulation of locomotor rhythms in the spinalized cat preparation; however, knee proprioceptors may be more important in human gait because of the differences in bipedal and quadrapedal locomotion and the need for an enhanced support mechanism at this joint. In human SCI, locomotor activity can be produced with "hip-only" walking, with the knees held in an extended posture (Dietz et al. 2002), suggesting that knee proprioceptors might not play a critical role in regulating locomotion. It is unknown whether hip afferents influence locomotor activity with the knees held in a more flexed posture.
Unlike knee proprioceptors, limb load afferents have proven to be an important modulator in the spinal centers for locomotion. Spinalized cat preparations have shown that swing phase will not initiate if the contralateral limb is not in position to bear load (Grillner and Rossignol 1978). Also, stepping infant studies have shown that the load on the ipsilateral limb must be decreased to initiate swing phase (Pang and Yang 2000). Chronic spinal cats also show the ability to correct for loss of ground support in the ipsilateral limb during treadmill walking. The cat achieves this correction by flexing the ipsilateral limb on stepping into a hole in the treadmill. The contralateral leg extends the knee and hip to help support the body until the ipsilateral limb is ready for weight support (Hiebert and Gorassini 1994). It was concluded that the lack of input from load sensitive afferents in the ipsilateral limb had a major contributing role in this response.
Similar evidence linking ankle load afferents to locomotor modulation has been found in the chronic human SCI. Limb loading during body weight–supported treadmill training modulates EMG activity of ankle musculature according to the phases of the gait cycle (Dietz et al. 2002; Harkema et al. 1997). The unilateral hip extension movements applied in this study did not incorporate an ankle-loading paradigm. We postulate that an absence of the ankle afferent input into the organized interneuronal pathways may cause the difference in the phasing of VM activity in this study (Fig. 11). For example, in the spinalized cat, removal of contralateral limb support, similar to treadmill walking where the contralateral limb steps into a hole, inhibits swing phase in the ipsilateral limb and even produces knee extensor activity (Hiebert and Gorassini 1994). The loading of the ankle at the appropriate time in the sinusoidal cycle would likely affect the phasing of the ankle and quadriceps musculature and needs to be examined further.
Effects of intersubject variability
Previous studies have also postulated that the level of the spinal cord lesion has an effect on the response (Dietz et al. 1999). This may be due, in part, to involvement of the cervical spinal regions in the locomotor generating circuits (Dietz 2002). No conclusions could be drawn about level of injury in this study because consistent responses were seen in subjects 5 (T6 ASIA A) and 1 (T6 ASIA A) and small responses were found in subjects 3 (C7–T1 ASIA B) and 8 (C4-5 ASIA C). Similarly, no conclusions about whether the location of the lesion or the nature of injury (incomplete vs. complete) had an effect on the response could be made because of the varying amounts of antispasm medications that each subject was taking at the time of the study. No significant differences in the magnitude and phase of the torques or EMGs were observed when comparing subjects on spasticity medications to those without medications (P > 0.10), although the sample size was small (n = 4, 5) for this comparison. Despite the lack of significant differences, the subjects with the largest and most consistent responses were not prescribed antispasm medications (subjects 6, 7, and 10); however, subject 9 yielded a consistent reflex response and this subject was taking baclofen. The effects of spasticity medications on the magnitude and phasing of muscle activity responses to hip oscillations is an important topic for future studies since the response to hip oscillation may be indicative of the clinical incidence of extensor spasms and may impact locomotor training.
Conclusions
This study has shown that hip afferent input can modulate the timing of lower extremity muscles in a predictable fashion, but hip proprioceptors alone cannot create consistent locomotor rhythms. The reflex response manifested from sinusoidal oscillations of the hip in subjects with chronic SCI is likely due to organized pathways located within the lumbosacral region of the spinal cord. We believe that these pathways are associated with the spinal centers for locomotion because of the organized timing of flexor and extensor muscles. The alternating flexor and extensor activity was not entirely consistent with locomotor patterns because of the ill-timed uniarticular knee extensor activity, which may be due, at least in part, to the absence of other afferent feedback such as limb load afferents. Locomotor training is becoming a very important tool in the rehabilitation protocol in human SCI. Proper afferent input must be present during this training to obtain appropriate muscle activation synergies.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: B. D. Schmit, Dept. of Biomedical Engineering, Marquette Univ., PO Box 1881, Milwaukee, WI 53201-1881 (E-mail: brian.schmit{at}marquette.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Andersson O and Grillner S. Peripheral control of the cat's step cycle: entrainment of the central pattern generators for locomotion by sinusoidal hip movements during "fictive locomotion." Acta Physiol Scand 118: 229–239, 1983.[ISI][Medline]
Ashworth B. Preliminary trial of carisoprodol in multiple sclerosis. Practioner 192: 540–542, 1964.
Batschelet E. Circular Statistics in Biology. New York: Academic Press, 1981.
Beres-Jones JA, Johnson TD, and Harkema SJ. Clonus after human spinal cord injury cannot be attributed solely to recurrent muscle-tendon stretch. Exp Brain Res 149: 222–236, 2003.[ISI][Medline]
Calancie B, Needham-Shropshire B, Jacobs P, Willer K, Zych G, and Green BA. Involuntary stepping after chronic spinal cord injury: evidence for a central rhythm generator for locomotion in man. Brain 117: 1143–1159, 1994.
Conway BA, Hultborn H, and Kiehn O. Proprioceptive input resets central locomotor rhythm in the spinal cat. Exp Brain Res 68: 643–656, 1987.[ISI][Medline]
Dietz V. Do human bipeds use quadrapedal coordination? Trends Neurosci 25: 462–467, 2002.[CrossRef][ISI][Medline]
Dietz V, Colombo G, Jensen L, and Baumgartner L. Locomotor capacity of spinal cord in paraplegic patients. Ann Neurol 37: 574–582, 1995.[CrossRef][ISI][Medline]
Dietz V, Muller R, and Colombo G. Locomotor activity in spinal man: significance of afferent input from joint and load receptors. Brain 125: 2626–2634, 2002.
Dietz V, Nakazawa K, Wirz M, and Erni T. Level of spinal cord lesion determines locomotor activity in spinal man. Exp Brain Res 128: 405–409, 1999.[CrossRef][ISI][Medline]
Dietz V, Wirz M, Curt A, and Colombo G. Locomotor pattern in paraplegic patients: training effects recovery of spinal cord function. Spinal Cord 36: 380–390, 1998.[CrossRef][ISI][Medline]
Dimitrejevic MR and Nathan PW. Spasticity in man. 1. Some features of spasticity. Brain 90: 1–30, 1967a.
Dimitrejevic MR and Nathan PW. Studies of spasticity in man. 2. Analysis of stretch reflexes in spasticity. Brain 91: 333–358, 1967b.
Dimitrejevic MR and Nathan PW. Studies of spasticity in man. 3. Analysis of reflex activity evoked by noxious cutaneous stimulus. Brain 91: 349–368, 1968.
Dobkin BH, Harkema SJ, Requejo P, and Edgerton VR. Modulation of locomotor-like EMG activity in subjects with complete and incomplete spinal cord injury. J Neurol Rehab 9: 183–190, 1995.[ISI][Medline]
Engsberg JR, Ross SA, Olree KS, and Park TS. Ankle spasticity and strength in children with spastic diplegic cerebral palsy. Devices Med Child Neurol 42: 42–47, 2000.[CrossRef]
Grillner S and Rossignol S. On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Res 146: 269–277, 1978.[CrossRef][ISI][Medline]
Harkema SJ. Neural plasticity after human spinal cord injury: applications of locomotor training to the rehabilitation of walking. Prog Clin Neurosci 7: 455–468, 2001.
Harkema SJ, Hurley SL, Patel UK, Requejo PS, Dobkin BH, and Edgerton VR. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol 77: 797–811, 1997.
Hiebert GW and Gorassini M. Corrective responses to loss of ground support during walking. II: Comparison of intact and chronic spinal cats. J Neurophysiol 71: 611–622, 1994.
Hiebert GW, Whelan PJ, Prochazka A, and Pearson KG. Contribution of hind limb flexor muscle afferents to the timing of phase transitions in the cat step cycle. J Neurophysiol 75: 1126–1136, 1996.
Kriellaars DJ, Brownstone RM, Noga BR, and Jordan LM. Mechanical entrainment of fictive locomotion in the decerebrate cat. J Neurophysiol 71: 2074–2086, 1994.
Kuhn R. Functional capacity of the isolated human spinal cord. Brain 73: 1–51, 1950.
Lam T and Pearson KG. Proprioceptive modulation of hip flexor activity during swing phase of locomotion in decerebrate cats. J Neurophysiol 86: 1321–1332, 2001.
Lance JW. Symposium synopsis. In: Spasticity: Disordered Motor Control edited by Feldman RG, Young RR, and Koella WP. Chicago, IL: Yearbook Medical Publishers, 1980, p. 185–203.
Little JW, Mickelson P, Umlauf R, and Britell C. Lower extremity manifestations of spasticity in chronic spinal cord injury. Am J Phys Med Rehab 68: 26–32, 1989.[CrossRef][ISI][Medline]
Pang M and Yang JF. The initiation of the swing phase in human infantstepping: importance of hip position and leg loading. J Physiol 528: 389–404, 2000.
Pepin A, Ladouceur M, and Barbeau H. Treadmill walking in incomplete spinal-cord-injured subjects: 2. Factors limiting the maximal speed. Spinal Cord 41: 271–279, 2003.[CrossRef][ISI][Medline]
Pang M and Yang JF. Interlimb co-coordination in human infant stepping. J Physiol 533: 617–625, 2001.
Perry J. Section 2: normal gait. Section 3: pathological gait. In: Gait Analysis: Normal and Pathological Function. NJ: SLACK, 1992, p. 49–347.
Schmit BD. Mechanical measures of spasticity in stroke:. Topics Stroke Rehab 8: 13–26, 2001.
Schmit BD and Benz EN. Extensor reflexes in human spinal cord injury: activation by hip proprioceptors. Exp Brain Res 145: 520–527, 2002.[CrossRef][ISI][Medline]
Schmit BD, McKenna-Cole A, and Rymer WZ. Flexor reflexes in chronic spinal cord injury triggered by imposed ankle rotation. Muscle Nerve 23: 793–803, 2000.
